were investigated. The probable difference between the corrected read- 

 ings of a pair of thermometers varied among the different groups from 

 0.004° to 0.021° C, and was .010° C. for the entire 1,432 comparisons. 

 The observed temperatures are therefore considered to be accurate to 

 0.01° C. 



The precision of the sahnity bridge measurements is regarded as 

 about ± 0.004 °/oo sahnity, although 38 double measurements of 

 Copenhagen standard w'ater, batch P15, resulted in ±0.002 °/oo salinity. 

 The accuracy of the measurements, however, is no better than the 

 titration method used in calibration of the salinit}' bridge. As very little 

 information has been published regarding the use of substandard 

 seawater, figure 1 is shown to illustrate the degree of variability under 

 conditions existing on board the General Greene. The points represent 

 the variation in salinity of substandard water as given from salinity 

 bridge comparisons with Copenhagen standard water of the batch Pi 5. 

 The substandard water was actual untreated sea water collected from 

 the surface and stored in oil-sealed carboys. During 1941, two carboys 

 were used, the first for the first and second surveys and the second for 

 the third survey and the post-season cruise. After completion of the 

 second survey, the remainder of the first carboy was transferred to 

 well-aged, selected citrate of magnesia bottles. These were subsequently 

 compared with Copenhagen water at the same time that the second 

 carboy comparisons were being made during the third survey and the 

 postseason cruise. It will be noted that in addition to a larger change 

 in salinity, the water stored in the citrate bottles showed about twice 

 the variability shown by the water stored in the carboys. Experience 

 with preliminary silver nitrate titrations indicates that during about 

 the first week after collection the salinity of water stored in an oil- 

 sealed carboy may shift by several hundredths of a part per thousand. 

 After that, more stable conditions seem to be established and the water 

 undergoes a gradual increase in salinity amounting to about 0.002 °/oo 

 per month, as shown in figure 1. Insufficient information is at hand to 

 permit description of the changes which take place beyond a period of 

 about 2 to 3 months, since on the General Greene it is usually necessary 

 to collect another carboy of substandard water before the end of about 

 3 months. 



The current maps prepared during the season as well as that showing 

 the current conditions existing during the postseason cruise were made 

 following the ])rocedure used during the past several >ears. In the 

 construction of the current charts showing conditions around the ice- 

 berg, a slightly different procedure was used. Whereas our practice has 

 bp'^n to compute densities (values of at) from temperature and salinit>- 

 values scaled from vertical distribution curves at the nominal depths 

 used in the numerical integration, in the treatment of the observations 

 made in the neighborhood of the iceberg densities were scaled from a 

 vertical distribution curve constructed with values computed from 



